Weaknesses in Integrating Welding Systems with Robots

“What is expected of a welding inverter” in order to be able to interact with a robot?  Common, logical questions for a welding equipment marketing guy, right? In a recent robotic welding group forum, Mr. Chinoy, the marketing manager of a welding equipment company, also asked “what parameters are required to integrate GMAW (MIG) equipment to the robot control panel”, besides wire feed speed and voltage? I answered those. And yet, hidden under the tip of that question like the 90% of an underwater iceberg, is the real question of ship-sinking power: what welding system interfacing and content will really earn the respect and repeat business of an end-user customer? Let’s do something stunning, and talk about that far bigger question too!

The answers depend on both your target business segment, and your company’s long-term goals as a welding equipment manufacturer. Many welding equipment companies design and launch a new machine every 3 years.  They just answer the obvious visible/functional questions.  One company chose to put out a pulse-MIG inverter that has been in continuous production for over 20 yrs, and has been the “king” not only of Electric Boat but the Korean shipyards for over a decade. In fact, older system versions can typically be upgraded to latest performance or customized waveform combinations with a simple plug-in EPROM chip swapout.

Why such content & success with the Digipulse (Automatic) system?  Simple – they answered the big hidden questions, and applied the hard yet hidden expertise required in order to faithfully serve the arc physics as well as the customer’s real needs and desires. How could they design that content back when welding robots were nearly non-existent? Because “hard-tooled” PLC-interfaced welding automation has essentially the same basic performance and interfacing needs as a robot. That can be shown by taking the unusual step of putting a robotic MIG process “fishbone” diagram together.

I’ve put a GMAW fishbone below (a W.E./SSBB project collaboration). It’s still hard to read when you click on it, and it doesn’t touch on the welding system design or integration content. But, it does show the overall process complexity and provides a starting point to consider welding system design and integration needs in order to consistently deliver perfect weld quality.

GMAW Process Fishbone

GMAW Process Automated Welding Fishbone

Take arc-starting and arc-established signals, for example. A manual welder is going to automatically compensate for an occasional poor arc-start. It doesn’t matter. But consider the dramatic difference in welding automation: when the torch travel in automation must rely on a signal to begin, aren’t the quality and cost implications much more dramatic and far-reaching than what most welding equipment manufacturers have been prepared to admit? This is only one piece of the automation puzzle, but it’s both critical and badly neglected.

The common minimalist approach is to provide a feedback signal during active welding, a “system ready” signal, an error output signal to indicate the system is in a fault state and unable to weld, and maybe a system-reset input. Many welding systems just provide those minimums, as “add-on” content to enable manual welding systems to go on a robot.

The problem from there is that many end users expect (at least eventually) to get high-performance welding automation results.  Of course that doesn’t happen, then everyone points a finger of blame at someone else, and if the customer succeeds in identifying the true welding-system design and/or integration weakness using many examples and actual real-time recorded data, the company responsible (such as Panasonic did, twice) might simply shrug and say “it welds good most of the time”.  Of course it does. But it’s also incapable of delivering world-class performance, simply because it’s not designed to.  

For another example, Miller, at least as much as other welding equipment companies, has a long history of their marketing department driving and making too many new equipment content decisions as if they are designing and selling Sony Playstations. As a result, many good welding process engineers who have been in the industry for 15 yrs or more, feel that Miller MIG equipment is one to stay away from if you need a sustainable platform that actually performs well and doesn’t prove to be an unsupported pain-in-the-butt in 3 years that isn’t actually compatible with previous versions. Of course, they’re mostly too polite to say it… they just try to avoid the equipment.

I applaud Mr. Chinoy for asking the original questions. And I suggest that my SWE readers share in comments how they would answer those questions. Yet, I concede that I didn’t answer them in specifics for high-performance robotic automation.  Why not?

My experience has been that unless a company is willing to hear and deal with those fundamental business philosophy issues, they won’t hear valuable technical answers no matter how much of my time I waste trying to tell them. (And I’ve wasted a LOT of effort, which has too often led to plotting by the famous pig to stab me in the back rather than learn to sing.) There are huge gains to be had in manufacturing welding, from selecting and properly implementing optimized high-performance welding systems. Unfortunately, too many welding OEM companies are more interested in marketing the illusions and assumptions of performance (“bells and whistles”), than they are interested in marketing their hard work invested in putting excellent foundational content into their hardware and software.

When the real problem “fix” is putting genuine basic expertise into the software, isn’t it customer abuse to say “we fixed that in our new system – just upgrade”?

There’s something about a “shut up and buy it” marketing approach that rubs me the wrong way, especially if I’m trying to take a Design For Six Sigma (DFSS) or DFM approach. Some assume that, like DFM, DFSS is a product-design thrust, but it’s also a process-design thrust that goes all the way to the “C” in DMAIC. I’ve been in countless “8-D” and Continuous Improvement teams trying to tackle welding defect reductions and root causes. I promise you that it’s difficult and frustrating as a welding engineer to concretely prove that welding system design deficiencies are the true root causes of your #1 largest welding defect type. But even more frustrating is after you have done that, and having production full of over 100 of that vendor’s systems, to have them refuse to put you in touch with their software designer to explain to them how to resolve the problems.

Some welding equipment enables a high degree of stable process Control, and OEM welding systems always claim to, but most don’t deliver.  At minimum, I want to partner with a vendor design group who sees collaboration as a superb C.I. opportunity to improve performance and demonstrate a commitment to help me deliver better quality and profitability on the production floor.

Imagine being ready to explain – for free – how your main GMAW equipment OEM can leap ahead in their system performance AND help you dramatically reduce your biggest welding defects at the same time… except they aren’t interested!

Or, imagine explaining to another legendary welding equipment company that you have two different scopes both proving that if you activate the weld from the remote terminal strip (like a robot or PLC, instead of with their manual Test Weld button), that the arc-established-signal output immediately comes on with the wire-drive, before the wire ever touches the part… but the only response you get is “that’s impossible”.  The 10-minute schematic analysis of a former Air Force electronics expert is that it’s not impossible: it means they have a purchased PCB component that is defective, in a non-typical way. It’s ironic that my expert could identify that, but the welding system OEM never could, and saw no need to modify their factory testing procedure to activate the terminal strip connection – even though every single piece of customer equipment would make every weld that way.
See the problem? That equipment performance guarantees weld defects (including short and missing welds) which are unlikely to be found in visual inspection.

For those managers wondering: yes, a SWE (Smart Welding Engineer) really can test and identify superior welding performance, rather than buying a bunch of new equipment from the best sales brochures while you cross your fingers again.

For those Smart Welding Engineers who are wondering, a “deep-memory” oscilloscope (like Agilent or HP) and an Arc Data Monitor (Impact Engineering) are valuable tools to examine actual welding system performance in critical technical areas such as arc starting, arc-established signal, and repeatable stable accuracy of arc voltage, current, and wire-feed-speed.

SWE Warnings:
Caution your boss that a data-driven approach puts a target on your back both internally and with vendors, and even your boss’ boss may get played as a patsy in a high-stakes industrial welding “survivor” game of outwit, outplay, outlast!  You might also want to ask him if, after putting in lots of work and money to identify the best approach, whether purchasing is going to be allowed to dictate a purchasing standard of inferior cheaper equipment “to save money” because the cost-justifications don’t meet their accounting criteria.

Welding processes are too complex to achieve world-class welding stability and performance levels in demanding high-volume automation, without carefully studying the physics and production realities, and then designing systems to achieve that level of performance. For the most part, robots are capable of delivering.  For the most part, welding systems are not.  Any content that is not skillfully designed to be there… won’t be.  Surprise!

Brian Dobben

Advertisements

2 Responses to Weaknesses in Integrating Welding Systems with Robots

  1. matt morris says:

    Finding customer specific requirements has long been a challenge. In fact, a survey of current automotive suppliers found that a significant number did not know where to go for the latest applicable customer specific requirements. For registrars and end users, this represents a serious problem. How can rules be followed and enforced if they are not readily available?

    It is with these challenges in mind customerspecifics.com was created. Here you will find a community to access, share, and discuss customer specific requirements.

    Please understand that the content of this site will take some time to develop as we work hand in hand with each of you to build a comprehensive database of the thousands of available customer specific requirements. We ask that you join us in our cause as we attempt to build something great for the common benefit of the quality community.

  2. […] Welding An Exceptional Welding Engineering Blog « Weaknesses in Integrating Welding Systems with Robots Turnkey Illusions – How to Avoid Pitfalls When Outsourcing Welding Automation April […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s

%d bloggers like this: